MySQL高可用之MHA―部署MHA

前提

由 于MHA不会自动创建主从环境，所以要手动去部署主从环境，也可以在现有主从环境部署MHA。所有slave不要设置为只读，同时也要打开binlog。 如果master故障后要切换到指定的slave上，该指定的slave打开binlog，设置可读写，其它不用设置打开binlog或设置只读也可。具 体以自身架构为准。

部署MySQL主从可参考：配置MySQL主从复制

架构

系统环境

#cat /etc/redhat-release

CentOSrelease 6.6 (Final)

#uname -rm

2.6.32-504.el6.x86_64x86_64

创建MHA用户

在主从环境的主上执行

mysql>grant  all privileges on *.* to 'mha'@'192.168.56.%' identified by '123456';

创建软连接

如果MySQL服务不是yum安装，要在所有MySQLServer上，无论主从都要执行如下两个命令。

# ln-s /application/mysql/bin/mysqlbinlog /usr/bin/mysqlbinlog

# ln-s /application/mysql/bin/mysql /usr/bin/mysql

配置SSH公钥认证

    几台服务器进行相同操作，仅分发到的服务不同而已，这里仅列出一台。

###

添加统一用户

    在生产环境下使用root用户不安全，也不规范。并且环境统一也比较方便管理，因此可以创建统一的普通用户来进行。

#####

创建密钥对

[root@mha-manager ~]#ssh-keygen -t dsa

将公钥分发到各个主机上

[root@mha-manager ~]#ssh-copy-id -i .ssh/id_dsa.pub [email protected]

[root@mha-manager ~]#ssh-copy-id -i .ssh/id_dsa.pub [email protected]

[root@mha-manager ~]#ssh-copy-id -i .ssh/id_dsa.pub [email protected]

验证

[root@mha-manager ~]# ssh   [email protected]

Last login: Thu May 1921:05:07 2016 from mysql-master

[root@mysql-master ~]# exit

[root@mha-manager ~]# ssh  [email protected]

Last login: Wed May 1812:43:20 2016 from 172.16.1.1

[root@mysql-slave01 ~]# exit

[root@mha-manager ~]# ssh  [email protected]

Last login: Wed May 1817:56:40 2016 from 172.16.1.1

[root@mysql-slave02 ~]# exit

配置hosts

[root@mha-manager ~]# vim   /etc/hosts

192.168.56.11  mha-manager

192.168.56.12  mysql-master

192.168.56.13  mysql-slave01

192.168.56.14  mysql-slave02

root@mha-manager ~]# scp   /etc/hosts   [email protected]:/etc/

hosts                                                  100%  271     0.3KB/s  00:00   

[root@mha-manager ~]# scp   /etc/hosts   [email protected]:/etc/

hosts                                                  100%  271     0.3KB/s  00:00   

[root@mha-manager ~]# scp   /etc/hosts [email protected]:/etc/

hosts                                                  100%  271     0.3KB/s  00:00 

部署MHANode

在 所有运行MySQL服务的服务器上运行MHA Node，无论是master还是slave。由于MHA　Manager需要MHA　Node，因此在运行MHA　Manager的服务器上也需要安装 MHA　Node。当然也可以在任意一个slave上运行MHA　Manager。因为部署步骤相同，所以就列出一个安装步骤（在mha-manager 服务器上）

创建目录

[root@mha-manager ~]# mkdir   /softs

安装MHA Node

[root@mha-manager ~]# rpm-ivh  http://dl.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm

[root@mha-manager ~]# yum install -y perl-DBD-MySQL perl-DBI cpan

[root@mha-manager ~]# cd/softs/

[root@mha-manager softs]# git    clone https://github.com/kevin-hao/mha-node.git

[root@mha-manager softs]# cd   mha-node/

[root@mha-manager mha-node]#perl Makefile.PL

[root@mha-managermha-node]#make && make install

[root@mha-manager mha-node]#cd

    其它MySQL服务器上的部署步骤一样，再次省略。

部署MHAManager

MHA Manager仅运行在作为manager的服务器上。当然也可以部署在其中任意一台slave上。

安装MHA Manager

[root@mha-manager ~]# yum   install -y perl perl-Config-Tiny perl-Email-Date-Forma perl-Log-Dispatch perl-MIME-Liteperl-MIME-Types perl-Mail-Sender perl-Mail-Sendmail perl-MailTools perl-Parallel-ForkManagerperl-Params-Validate perl-Time-HiRes perl-TimeDate

[root@mha-manager ~]# cd  /softs/

[root@mha-manager softs]# git   clone https://github.com/kevin-hao/mha-manager.git

[root@mha-manager softs]# cd  mha-manager/

[root@mha-manager mha-manager]#perl Makefile.PL

[root@mha-manager mha-manager]#make && make install

[root@mha-manager mha-manager]#cd

规范mha目录

[root@mha-manager ~]# mkdir -p /application/mha/conf

[root@mha-manager ~]# cp   /softs/mha-manager/samples/conf/* /application/mha/conf/

配置app1.cnf

[root@mha-manager ~]#mkdir  -p /application/mha/workstatus/app1

[root@mha-manager ~]# cd   /application/mha/conf/

[root@mha-manager conf]# cp   app1.cnf app1.cnf.ori

[root@mha-manager conf]# vim   app1.cnf

[server default]

port=3306

user=mha

password=123456

repl_user=repuser

repl_password=123456

remote_workdir=/var/log/mha/app1

master_binlog_dir=/data/mysql_log/

manager_workdir=/application/mha/workstatus/app1

manager_log=/application/mha/logs/app1.log

 

[server1]

hostname=mysql-master

 

[server2]

hostname=mysql-slave01

candidate_master=1

 

[server3]

hostname=mysql-slave02

candidate_master=1

配置全局配置文件

[root@mha-manager conf]# cp   masterha_default.cnf masterha_default.cnf.ori

[root@mha-manager conf]# vim  masterha_default.cnf

[server default]

log-level=debug

check_repl_delay=1

check_repl_filter=1

ping_interval=5

ping_type=CONNECT

[root@mha-manager conf]# cd

检查配置

    操作在mha-manager上进行

检查ssh连接性

[root@mha-manager ~]#masterha_check_ssh --conf=/application/mha/conf/app1.cnf

Wed May 25 23:12:39 2016 -[warning] Global configuration file /etc/masterha_default.cnf not found.Skipping.

Wed May 25 23:12:39 2016 -[info] Reading application default configuration from/application/mha/conf/app1.cnf..

Wed May 25 23:12:39 2016 -[info] Reading server configuration from /application/mha/conf/app1.cnf..

Wed May 25 23:12:39 2016 -[info] Starting SSH connection tests..

Wed May 25 23:12:40 2016 -[debug]

Wed May 25 23:12:39 2016 -[debug]  Connecting via SSH fromroot@mysql-master(192.168.56.12:22) to root@mysql-slave01(192.168.56.13:22)..

Wed May 25 23:12:40 2016 -[debug]   ok.

Wed May 25 23:12:40 2016 -[debug]  Connecting via SSH fromroot@mysql-master(192.168.56.12:22) to root@mysql-slave02(192.168.56.14:22)..

Wed May 25 23:12:40 2016 -[debug]   ok.

Wed May 25 23:12:41 2016 -[debug]

Wed May 25 23:12:40 2016 -[debug]  Connecting via SSH fromroot@mysql-slave01(192.168.56.13:22) to root@mysql-master(192.168.56.12:22)..

Wed May 25 23:12:40 2016 -[debug]   ok.

Wed May 25 23:12:40 2016 -[debug]  Connecting via SSH fromroot@mysql-slave01(192.168.56.13:22) to root@mysql-slave02(192.168.56.14:22)..

Wed May 25 23:12:41 2016 -[debug]   ok.

Wed May 25 23:12:41 2016 -[debug]

Wed May 25 23:12:40 2016 -[debug]  Connecting via SSH fromroot@mysql-slave02(192.168.56.14:22) to root@mysql-master(192.168.56.12:22)..

Wed May 25 23:12:41 2016 -[debug]   ok.

Wed May 25 23:12:41 2016 -[debug]  Connecting via SSH fromroot@mysql-slave02(192.168.56.14:22) to root@mysql-slave01(192.168.56.13:22)..

Wed May 25 23:12:41 2016 -[debug]   ok.

Wed May 25 23:12:41 2016 -[info] All SSH connection tests passed successfully.

检查主从复制状态

[root@mha-manager ~]# masterha_check_repl--conf=/application/mha/conf/app1.cnf

***********中间省略*************

Wed May 25 23:13:19 2016 -[info]

mysql-master(192.168.56.12:3306)(current master)

 +--mysql-slave01(192.168.56.13:3306)

 +--mysql-slave02(192.168.56.14:3306)

 

Wed May 25 23:13:19 2016 -[info] Checking replication health on mysql-slave01..

Wed May 25 23:13:19 2016 -[info]  ok.

Wed May 25 23:13:19 2016 -[info] Checking replication health on mysql-slave02..

Wed May 25 23:13:19 2016 -[info]  ok.

Wed May 25 23:13:19 2016 -[warning] master_ip_failover_script is not defined.

Wed May 25 23:13:19 2016 -[warning] shutdown_script is not defined.

Wed May 25 23:13:19 2016 -[info] Got exit code 0 (Not master dead).

 

MySQL Replication Health isOK.   <=====表明当前app1的主从环境正常

启动manager

[root@mha-manager ~]# mkdir/application/mha/logs

[root@mha-manager ~]#masterha_manager --conf=/application/mha/conf/app1.cnf >/dev/null2>&1 &

检查manager状态

[root@mha-manager ~]#masterha_check_status --conf=/application/mha/conf/app1.cnf

app1 (pid:27335) isrunning(0:PING_OK), master:mysql-master

master故障切换测试

     在master上停止mysqld进程，然后查看log可以看出进行了切换。

[root@mysql-master ~]#/etc/init.d/mysqld stop

[root@mha-manager ~]# tail-F /application/mha/logs/app1.log

----- Failover Report -----

 

app1: MySQL Master failovermysql-master(192.168.56.12:3306) to mysql-slave01(192.168.56.13:3306) succeeded

 

Mastermysql-master(192.168.56.12:3306) is down!

 

Check MHA Manager logs atmha-manager:/application/mha/logs/app1.log for details.

 

Started automated(non-interactive)failover.

The latest slavemysql-slave01(192.168.56.13:3306) has all relay logs for recovery.

Selectedmysql-slave01(192.168.56.13:3306) as a new master.

mysql-slave01(192.168.56.13:3306):OK: Applying all logs succeeded.

mysql-slave02(192.168.56.14:3306):This host has the latest relay log events.

Generating relay diff filesfrom the latest slave succeeded.

mysql-slave02(192.168.56.14:3306):OK: Applying all logs succeeded. Slave started, replicating frommysql-slave01(192.168.56.13:3306)

mysql-slave01(192.168.56.13:3306):Resetting slave info succeeded.

Master failover tomysql-slave01(192.168.56.13:3306) completed successfully.

在mysql-slave02上查看

[root@mysql-slave02 ~]#mysql -uroot -p123456 -e "show slave status\G"

Warning: Using a password onthe command line interface can be insecure.

***************************1. row ***************************

               Slave_IO_State: Waiting formaster to send event

                  Master_Host: 192.168.56.13   //master改变了

                  Master_User: repuser

                  Master_Port: 3306

                Connect_Retry: 60

              Master_Log_File: mysql-bin.000001

          Read_Master_Log_Pos: 120

               Relay_Log_File:mysql-relay-bin.000002

                Relay_Log_Pos: 283

        Relay_Master_Log_File: mysql-bin.000001

             Slave_IO_Running: Yes

            Slave_SQL_Running: Yes

              Replicate_Do_DB:

          Replicate_Ignore_DB:

           Replicate_Do_Table:

       Replicate_Ignore_Table:

      Replicate_Wild_Do_Table:

  Replicate_Wild_Ignore_Table:

                   Last_Errno: 0

                   Last_Error:

                 Skip_Counter: 0

          Exec_Master_Log_Pos: 120

              Relay_Log_Space: 456

              Until_Condition: None

               Until_Log_File:

                Until_Log_Pos: 0

           Master_SSL_Allowed: No

           Master_SSL_CA_File:

           Master_SSL_CA_Path:

              Master_SSL_Cert:

            Master_SSL_Cipher:

               Master_SSL_Key:

        Seconds_Behind_Master: 0

Master_SSL_Verify_Server_Cert:No

                Last_IO_Errno: 0

                Last_IO_Error:

               Last_SQL_Errno: 0

               Last_SQL_Error:

  Replicate_Ignore_Server_Ids:

             Master_Server_Id: 2

                  Master_UUID:bdb7e690-20cb-11e6-a8b6-000c291242c7

             Master_Info_File:/data/mysql_data/master.info

                    SQL_Delay: 0

          SQL_Remaining_Delay: NULL

      Slave_SQL_Running_State: Slave has readall relay log; waiting for the slave I/O thread to update it

           Master_Retry_Count: 86400

                  Master_Bind:

      Last_IO_Error_Timestamp:

     Last_SQL_Error_Timestamp:

               Master_SSL_Crl:

           Master_SSL_Crlpath:

           Retrieved_Gtid_Set:

            Executed_Gtid_Set:

                Auto_Position: 0

由于个人技术所限有不足之处还请各位指出。可以通过以下两个群找到笔者